Answer
The design was flawed.
Work Step by Step
The problem with the experiment is that although there can be $6$ degrees of separation, nothing guarantees that anyone will know whoe to choose to reach the targeted person. Also a more random sample that included a larger part of the US population should have been used.
